Both Old French meters and their Modern French descendants are usually thought to lack the internal binary constituent structure of, say, English or German iambic verse. In this article, however, an underlying iambic structure for the Old French octosyllable is established through quantitative analysis of a large corpus of texts written from c. 975 to 1180 (42 distinct works, including over 22,000 lines). Because no texts conform absolutely to the grammar of English iambic verse (Halle & Keyser, 1971; Kiparsky, 1977), certain measures are proposed for the degree to which a sample deviates from the iambic pattern; these values are then compared with the (chance) deviation of normal Old French prose. A significant correlation emerges between these measures and date of composition, author, and genre: early texts are almost perfectly iambic, and late 12th-century texts approach, but do not reach, chance levels. It is concluded that the grammar of meter used by Old French authors underwent a gradual change during the 12th century, a change comparable to more familiar phonological and syntactic changes.

Nigerian Arabic has two structures for coding N + N relations: one in which the properties of the possessed noun are severely constrained and one in which the possessed and possessor nouns allow a full complement of modifiers. Similar to the methodology of Poplack and her collaborators (e.g., Sankoff, Poplack, & Vanniarajan, 1990), a normative distribution of nouns in the two possessive structures is established based on a corpus of monolingual (non-codeswitched) Nigerian Arabic texts. In a corpus of codeswitched texts, the distribution of English lexical insertions is found to deviate markedly from these normative patterns. The notion of iconicity is invoked to explain the skewed insertional patterns. It is hypothesized on the basis of various psycholinguistic studies that insertions from English are harder to access than are native lexemes. To compensate for slower access time, speakers match these insertions with the possessive structure that requires a minimal amount of manipulation for rapid embedding. This, it is shown, is the less iconic of the two possessive constructions. What emerges is a distinctive pattern of possessives characteristic of codeswitched Nigerian Arabic. After briefly testing the iconicity hypothesis against insertional patterns in Hausa and Standard Arabic, two other languages well attested in the codeswitched corpus, the question is addressed as to whether the codeswitched variety represents a code unto itself.

Several changes in consonant and vowel pronunciations in younger generations of native speakers of Received Pronunciation (RP) are currently the object of research interest. In order to further an empirically grounded description of changes in RP, the present study examines variation in weak vowels. Patterns of variation in word-final open weak syllables (happy, city) as well as in past and present/plural suffixes (waited, changes) are investigated acoustically in the interview speech of eight young (born in the late 1970s) speakers of modern RP. The data show variation in happy vowels for some speakers according to phonetic environment, a phenomenon which deserves further study. kit/schwa variation in the inflectional suffixes studied here shows a tendency to maintain kit-like values. Overall, the study indicates that acoustic analysis of such weak vowels can provide interesting data on variation.

Hedges in Japanese conversation are discussed in relation to three social variables: age, sex, and formality. In the present study, hedges are defined as expressions of uncertainty, possibility, or tentativeness, all of which convey a sense of vagueness. I quantified the use of 26 Japanese hedges (e.g., toka ‘or something’, kana ‘I wonder’, nanka ‘like’, and tabun ‘probably’). Three hypotheses were posited to investigate the effect of the social variables: hedges are used (1) more often by younger speakers than by older speakers; (2) more often by women than by men; (3) more often in casual speech than in formal speech. Data of 20 single-sex dyads were collected from two age groups and from both sexes. The data supported hypotheses (1) and (2), but not (3). Most notably, younger female speakers use hedges often. Social factors of the speaker as well as context play an important role in the use of hedges in Japanese conversation.